My Photo

« BBC Infographics on Iraq | Main | NLP Search: On Entity Extractors »

February 13, 2007

Comments

p-air

BTW Matt, WRT:

"a system with a modicum of world knowledge would be off to a running start knowing that a person was being discussed (simple entity extraction would deal with this)"

Thought I'd experiement w/a decent entity extractor to see how it would deal w/the sentence "ian anderson was born in". I went to ClearForest's SWS Web interface (http://sws.clearforest.com/SWS.htm) and copied that sentence. While generally speaking on a larger corpus of text it performs pretty well in terms of finding proper names, on just 5 words it found no recognizable entities. When I put the sentence "Ian was born in 1947 in Dunfermline, Fife, Scotland", it found "Fife" to be a city and "Scotland" to be a country.

Interesting hunh? Tough to look too far down the road when this level of functionality is still far fm resolved adequately.

p-air

FYI, just tried the Named Entity extractor (http://www.alias-i.com:8080/lingpipe-demos/ne_en_news_muc6/textInput.html) on LingPipe w/the same sentence as I tried w/SWS and it too didn't pick up any entities there.

The comments to this entry are closed.

Twitter Updates

    follow me on Twitter

    March 2016

    Sun Mon Tue Wed Thu Fri Sat
        1 2 3 4 5
    6 7 8 9 10 11 12
    13 14 15 16 17 18 19
    20 21 22 23 24 25 26
    27 28 29 30 31    

    Categories

    Blog powered by Typepad